Object based image analysis fpr predictive substrate distribution map using multibeam echo sounder (MBES) and backscatter in Bidong Island

Abstract

This study conducted to map a seafloor substrate and the area that was covered was in South West of Bidong Island which located in Terengganu. The first part aims is to delineate meaningful pattern of substrate distribution in backscatter and multibeam bathymetry using Object Based Image Analysis (OBIA) and the second part classifies type of seafloor substrate using multibeam bathymetry and backscatter data. An acquisition of acoustic for multibeam bathymetry, backscatter mosaic and ground truth data in South West of Bi dong Island is used to characterize the sea floor substrate. Acoustic data of were collected R2Sonic 2020. For ground truth data, sample of sediment has been grab and classified based on its texture from triangle plot, thus classifying the sediment as coarse sand and very coarse sand. The data of mutibeam bathymetry, backscatter image and parameters from mutibeam were used for predicting the type of substrate that can be found on these seafloor. These combination of data were runs using Benthic Terrain Modeler (BTM), ISO Cluster unsupervised and Remote Sensing Object Based Image Analysis (RSOBIA).